Neural Face Identification in a 2D Wireframe Projection of a Manifold Object

Kehan Wang1     Jia Zheng2     Zihan Zhou2
1University of California, Berkeley     2Manycore Tech Inc.

faceformer identifies face loops from a single line drawing.

Abstract

In computer-aided design (CAD) systems, 2D line drawings are commonly used to illustrate 3D object designs. To reconstruct the 3D models depicted by a single 2D line drawing, an important key is finding the edge loops in the line drawing which correspond to the actual faces of the 3D object. In this paper, we approach the classical problem of face identification from a novel data-driven point of view. We cast it as a sequence generation problem: starting from an arbitrary edge, we adopt a variant of the popular Transformer model to predict the edges associated with the same face in a natural order. This allows us to avoid searching the space of all possible edge loops with various hand-crafted rules and heuristics as most existing methods do, deal with challenging cases such as curved surfaces and nested edge loops, and leverage additional cues such as face types. We further discuss how possibly imperfect predictions can be used for 3D object reconstruction.

Interactive Results

Input image
Ours
Ground truth

BibTeX

@inproceedings{faceformer,
  author    = {Wang, Kehan and Zheng, Jia and Zhou, Zihan},
  title     = {Neural Face Identification in a 2D Wireframe Projection of a Manifold Object},
  booktitle = {CVPR},
  pages     = {1622--1631},
  year      = {2022},
}

Acknowledgements

This work was done during Kehan Wang's internship at Manycore Tech Inc.